our public forum gets 5-10 spam registrations per day. We have Akismet enabled and trust level 0 restrictions but spam still gets through. What additional layers of protection work?
Layers that work well together:
- Block disposable email domains (there are Discourse settings for this)
- Require email confirmation
- Increase the minimum body length for first posts
- Enable the ‘approve unless trust level’ setting for TL0 posts
- Add a simple question to the registration form (not a CAPTCHA, just ‘What is this forum about?’)
- Block VPN/proxy registrations if your community does not need them
5-10 per day is manageable with manual moderation plus these settings. If it scales up, look at CleanTalk as an Akismet alternative.
on the server side, fail2ban with a custom rule for failed registration attempts helps. Also rate-limit the registration endpoint in your nginx config if you are reverse-proxying. Most spam bots give up after a few rate-limited attempts.
the registration question approach is low-tech but effective. Adding it along with the disposable email block. The VPN blocking is too aggressive for our audience since we have members in restricted countries.
Always negotiate the renewal terms before signing the initial contract. SaaS vendors discount year one and raise prices at renewal when switching costs are high.
This is helpful but overwhelming. Just tell me the simplest path that works for a small organization.